Role family
Analytics
Turn messy sales, campaign and customer data into decisions a CMO will actually act on
12
roles on campus
11
companies
What this role actually is
Analytics roles for marketing PGDMs sit between raw business data and the marketing decision: you measure what worked, segment who to target next, and forecast what happens if nothing changes. On our campus this family splits into three flavours. Analytics services firms (C5i, Cartesian Consulting) put you on global brand clients, doing project work for a fee. IT majors (Infosys, HCL) place you inside an analytics or consulting practice serving enterprise accounts. In-house roles (Piccadily Agro, Radhakrishna Foodland, Credresolve) hand you one business's sales, supply-chain or collections dashboards to own end to end. Euromonitor (syndicated market intelligence) and Arcesium (financial data operations) are the outliers, but they test the same two muscles every seat here tests: rigour with numbers, and the ability to explain those numbers to someone who will not read your workbook.
Day to day
- Pull and clean data in SQL and Excel from CRM, POS or campaign platforms; budget 30 to 40 percent of your week for wrangling and reconciliation, not modelling
- Build and maintain the weekly business dashboard (Power BI or Tableau) a client or category head lives on: sales versus target, campaign ROI, distributor fill rates
- Run customer segmentation (RFM, cohort, CLV) so the CRM team knows which segment gets which offer; this is the bread and butter at a Cartesian-type firm
- Measure campaign effectiveness through pre versus post lift, A/B test readouts and channel attribution, then defend the methodology when the client pushes back
- Write the 'so what' slide: compress a 40-tab workbook into five slides with one recommendation, then present it on the weekly client or leadership call
- In in-house roles (foods, agro, collections) forecast demand or recovery rates, flag the SKUs or accounts drifting off target, and chase other functions for the data behind the miss
Skills that matter
- SQL: joins, GROUP BY, and window functions; the single most-tested hard skill in these interviews and the one that decides most shortlists
- Advanced Excel: pivots, XLOOKUP or INDEX-MATCH, and what-if analysis; still the daily workhorse across Indian analytics teams
- One BI tool taken to portfolio depth: Power BI or Tableau, with at least one published dashboard you can actually demo on a call
- Statistics you can explain rather than just run: regression, significance testing, correlation versus causation, and sample bias
- Marketing measurement frameworks: CLV, RFM, funnel metrics, ROAS and CAC, and the basics of market mix; this is your edge over engineering-only candidates
- Structured problem-solving: breaking 'sales fell 12 percent, why?' into a MECE tree while the clock runs
- Data storytelling: turning an analysis into one line of insight plus a recommendation a non-analyst will buy
Who fits
The marketer who opens the dashboard before forming an opinion. You do not need an engineering background; most analysts at C5i or Cartesian come from commerce or BBA routes. What you do need is genuine tolerance for sitting with a dataset for hours, a pedantic streak about whether two numbers are even comparable, and the nerve to stand up and sell the finding once you have it. It suits people who found the Marketing Research and Analytics electives energising rather than a chore, and who would rather prove a campaign worked than write its copy. If ambiguous briefs frustrate you and you prefer questions with a right answer, this family will sit better with you than brand or sales roles.
What interviewers ask
Write a SQL query for the top 5 customers by revenue last quarter, or customers who bought in Jan but not Feb.
How to answer: Talk while you write: name the tables and keys you are assuming, use a clean GROUP BY with ORDER BY and LIMIT, and say which join type and why. If syntax escapes you, write correct pseudocode and flag it; interviewers at C5i and Infosys reward the logic over memorised syntax, and the Jan-not-Feb version is usually a LEFT JOIN with a NULL check or a NOT IN.
A client's e-commerce sales dropped 15 percent last month. How would you investigate?
How to answer: Ask two or three clarifying questions first (which category, which geography, is it revenue or units), then go MECE: internal (price, stock-outs, site changes, campaign pauses) versus external (competition, seasonality, category decline), and decompose into traffic times conversion times AOV. They are grading the structure, not the guess.
Explain RFM segmentation or CLV to a client who is not from analytics.
How to answer: Anchor on one concrete customer: someone who bought last week, buys monthly and spends Rs 5,000 a trip is a champion; someone silent for eight months needs a win-back offer. Then say what the business does differently for each. Jargon-free beats textbook-perfect here, and the follow-up is usually 'so what do we do about it'.
How do you know your campaign actually caused the sales lift, not just coincided with it?
How to answer: Explain test versus control, randomisation and statistical significance in plain words, then name one real trap: contamination between groups, seasonality, or a sample too small to trust. Saying 'correlation is not causation' out loud is usually the exact checkbox they are waiting for.
Guesstimate: how many bottles of whisky are sold in Haryana per month, or how many delivery trucks a QSR chain needs in Mumbai.
How to answer: Common at Piccadily and Foodland-type in-house roles. Go top-down (population, drinking-age share, drinkers, frequency, brand share), state every assumption aloud, keep the arithmetic round, and sanity-check the final number against something you know. They want a defensible method, not the 'correct' answer.
Walk me through a data project from your PGDM: what did you find, and what changed because of it?
How to answer: Pick one project (live project, summer internship, competition) and narrate it as problem, data, method, insight, action, with one real number attached. The 'what changed' is what separates you from the pack, so have it sharp and be ready for two follow-up 'why that method' questions.
What dashboards have you built in Tableau or Power BI, and what would you put on page one for a sales head?
How to answer: A link or screenshots beat any claim, so arrive demo-ready. For page one: four or five KPIs against target, a trend line, and one drill-down (region or SKU). Justify each tile by the decision it drives, not by how it looks.
How to prep
- 1. Get SQL to interview-fluency in three to four weeks: work through 50 to 70 easy and medium problems on a practice platform, concentrating on joins, GROUP BY with HAVING, and window functions. This single skill filters out half the marketing cohort in analytics shortlists.
- 2. Build one portfolio dashboard on real Indian data (an FMCG sales set, or your anonymised summer-internship data) in Power BI or Tableau, and rehearse a three-minute walkthrough. It converts 'knows the tool' from a claim into evidence you can show on the call.
- 3. Prepare two project stories in problem, data, method, insight, action format with actual numbers, then stress-test them: assume every C5i or Cartesian interviewer will ask 'why that method' and 'what would you do differently'.
- 4. Do 10 to 15 timed cases with a partner across the three shapes that recur here: sales-decline diagnosis, market sizing, and campaign ROI. Both services firms and in-house roles open with one, and structure under a clock is a trained reflex, not a talent.
- 5. Study each firm's actual business before its PPT: Cartesian is customer and CRM analytics, C5i serves global CPG and tech clients, Euromonitor sells syndicated research, Credresolve is collections. Pitching a segmentation idea at Euromonitor signals you skipped the homework.
Where it leads
Entry is typically Business Analyst or Analyst-Consultant at the standard PGDM analytics band. Over three to five years the paths fork. At services firms (C5i, Cartesian, the Infosys and HCL practices) you climb Analyst to Senior Analyst to Team Lead or Engagement Manager, owning a client relationship and a pod of three to five people, often jumping to a captive GCC or a product company for a 40 to 60 percent bump along the way. In-house analysts (agro, foods, collections) grow toward Category, Revenue or Sales-Ops Manager roles where analytics becomes the springboard into P&L ownership. The two premium exits are marketing-analytics leadership at a D2C or e-commerce firm (the Growth or CRM Head track) and a lateral into consulting; both reward people who kept one foot in business context rather than becoming pure tool operators. The compounding is real: SQL plus measurement plus storytelling stays scarce among marketers, which gives this family some of the strongest five-year optionality of any placement track on campus.